Black Lion Audio...

[jeremy.c.]

Originally Posted by madtheory

Yes, there are frontiers in astrophysics, improving the resolution of optical and radio telescopes, in pushing the limits of microprocessor manufacturing, stuff like that. However, S/Pdif communications are entirely artificial and well established, so there are no unknowns. Between transmitter, receiver and cable, the clock either works or it doesn't. Yes, there "may" be modulation (vague isn't it?). But in this case there are no subtleties. The receiver has an error margin, the clock gets to its destination, if not there's no audio. Sound quality is not affected. Don't take my word for it, read up on how it works and you'll see.

So to make a good S/Pdif cable, you need 75 ohm cable and solid connections. As I said, I have an ancient free cable and it's never caused any problems. I've sat there with a bit scope and checked! Easy to do, dead boring.

This is what I mean about establishing a base of what is taken for granted. If one knows a little about digital audio, the quote about clock modulation reveals a lot. Snake oil vendors throw in sciencey sounding terms, and are always vague. It take effort to interrogate them. Effort like reading and learning and stuff.

I'd recommend Ben Goldacre's book "Bad Science" to learn about good critical analysis of science, and pseudo science.

Well there's the rub isn't it?
The problem is the somewhat subjective nature of what sounds "good."
I had one of the first FW1814 mods and they even quoted me off this site on the 1814 page for a while. They took a fairly crap interface and definitely improved it, but not so much that I didn't sell it 6 months later. I felt that their claims, or what I perceived to be their claims, were oversold, and I see a lot of people over selling their products for them. But obviously this issue goes a lot deeper on this forum and a good chunk of pro-audio sales seems to come from specious claims. Ethan Winer is a great example of every audio marketer's nightmare, and you may agree with Ethan and his tactics or not, but he makes some very sound and salient points about what we are hearing vs. what we think we are hearing.
But it seems in any discussion like this one at the end of the day it really comes down to confirmation bias, buyer's remorse, expectation bias, peer pressure and on and on and discussion ad nauseam about subjective experiences. Unfortunately for the audiophiles (I noticed that there are a few on this thread keeping an "open" mind) and the believers in golden ears that extend beyond reasonable testing there does need to be a balance between subjective experience and rational questions.
I do think 90% of BLA's product is hype and incremental advances in quality, but this doesn't make them much different than any other audio manufacturer. You just get the added bonus of no resale value if the gear isn't living up to it's hype, because obviously, the only people who get to hear how this stuff actually sounds are the people who are already financially and psychologically invested in it. We would all do a service to ourselves if we could actually remain that open minded about things. Constantly questioning scientific understanding is only open minded if you're willing to go with the evidence when it presents itself.
So how you're getting a bad rap for being skeptical is beyond me unless viewed in the light of true believers, because I haven't seen the type of behavior you're being accused of.

Thank you for recommending Ben Goldacre's book, more people need to read stuff like that. Being scientifically literate and a rational thinker is more fun than it sounds!

edit: I should also point out that the guy I sold the FW1814 to sold it to someone else, and that guy sold it to yet another guy. They contacted me with questions at each step of the sale. So I have a confirmed 3 out of 4 people that were not satisfied with the quality of the mod

[zacheus83]

Originally Posted by Budda

As I've mentioned...

The issue with A/B'ing a mult'ed session with LE is that you need two separate computers running the two boxes. Unless you're an edit facility, do you have the ability to run two separate sessions? If so, would you have one "stock" unit of whatever and one "modded"?

I would like to do this test at some point. If I understand correctly, I would be testing the pre's by connecting each unit (modded and unmodded) to seprate computers and splitting a source signal between the 2... I assume we'd have to critique these in a blind test? Is there a signal we can run (like 1khz tone) where we can actually look at data as opposed to listening?

And then, what would be the proper way to test the difference in the D-A? Differences in the clock?

I have a buddy with a laptop (and I have an Intel mac pro), not sure if he has an 002R or 003R, but I will find an unmodded 002R and do the test. Point me in the direction of what would make the test unbiased. Thanks guys.

[madtheory]

jeremy.c, +1, I agree, well put.

Originally Posted by TurboJets

OK Einstein...how the he// am I going to do that if my box has already been modded.

If you like, I could do a loopback on my FW1814 using the file on the mastering thread I linked to earlier, and you could do the same on yours, and we'll post the results. How about it? Actions speak louder than words.

[Budda]

Originally Posted by zacheus83

I would like to do this test at some point. If I understand correctly, I would be testing the pre's by connecting each unit (modded and unmodded) to seprate computers and splitting a source signal between the 2... I assume we'd have to critique these in a blind test? Is there a signal we can run (like 1khz tone) where we can actually look at data as opposed to listening?

And then, what would be the proper way to test the difference in the D-A? Differences in the clock?

I have a buddy with a laptop (and I have an Intel mac pro), not sure if he has an 002R or 003R, but I will find an unmodded 002R and do the test. Point me in the direction of what would make the test unbiased. Thanks guys.

The only need for having the two computers is because the Digi products don't run in stand-alone mode. They won't even pass audio (I think) unless you've got a PT session open and the box configured. If you cared to level-match your inputs (smart!), then you can run a test tone from your single source (1kHz), split how you would split it (sometimes a split load or impedance "not matching" can create electronic differences that would account for audible differences), and adjust accordingly. From there, you don't need any sort of sync to match the sessions (make sure your bit depth and sample rates match!), but it would be good if you do a "slate" - clap your hands, tap the mic, something with a good clear initial transient for you to line up your recorded tracks to each other for the listening sessions. Then you can look at the pictures after (which "may" show you the difference), but you can also try listening on the stock and modded units (which will additionally show you the D/A difference).

I would additionally suggest that the difference will become much more apparent on stacking tracks, so you may wish to try tracking a song, even roughly performed, and do an "ITB" mix that can be identically set up between the two systems. Play them back to back on just one system (either one), and you will see the cumulative effect (if there is one, ha ha) of the modded input stages; if you play the same mix on both systems you can then also see how the stock and modded units D/A affects how you hear the same mix as far as reverbs, detail, clarity, etc.

Do I have it about right, naysayers? Loopback tests aside, of course - I'd rather listen to the effect on real audio than the atypical pure sine wave of a loopback test, and as I've said before, Neves would = MASSIVE FAIL on a loopback test.

[dandeurloo]

I think a single track test won't really show the benefit of the BLA mod either? Because clocking adds up the more tracks you use. Best would be to do the shot out of one session on a mod and un modded 002 or something.

I think with a test like that it will be night an day for you guys to hear.

[D K]

Originally Posted by jeremy.c.

Well there's the rub isn't it?
The problem is the somewhat subjective nature of what sounds "good."
I had one of the first FW1814 mods and they even quoted me off this site on the 1814 page for a while. They took a fairly crap interface and definitely improved it, but not so much that I didn't sell it 6 months later. I felt that their claims, or what I perceived to be their claims, were oversold, and I see a lot of people over selling their products for them. But obviously this issue goes a lot deeper on this forum and a good chunk of pro-audio sales seems to come from specious claims. Ethan Winer is a great example of every audio marketer's nightmare, and you may agree with Ethan and his tactics or not, but he makes some very sound and salient points about what we are hearing vs. what we think we are hearing.
But it seems in any discussion like this one at the end of the day it really comes down to confirmation bias, buyer's remorse, expectation bias, peer pressure and on and on and discussion ad nauseam about subjective experiences. Unfortunately for the audiophiles (I noticed that there are a few on this thread keeping an "open" mind) and the believers in golden ears that extend beyond reasonable testing there does need to be a balance between subjective experience and rational questions.
I do think 90% of BLA's product is hype and incremental advances in quality, but this doesn't make them much different than any other audio manufacturer. You just get the added bonus of no resale value if the gear isn't living up to it's hype, because obviously, the only people who get to hear how this stuff actually sounds are the people who are already financially and psychologically invested in it. We would all do a service to ourselves if we could actually remain that open minded about things. Constantly questioning scientific understanding is only open minded if you're willing to go with the evidence when it presents itself.
So how you're getting a bad rap for being skeptical is beyond me unless viewed in the light of true believers, because I haven't seen the type of behavior you're being accused of.

Thank you for recommending Ben Goldacre's book, more people need to read stuff like that. Being scientifically literate and a rational thinker is more fun than it sounds!

edit: I should also point out that the guy I sold the FW1814 to sold it to someone else, and that guy sold it to yet another guy. They contacted me with questions at each step of the sale. So I have a confirmed 3 out of 4 people that were not satisfied with the quality of the mod

There is a massive difference between your statements and madtheory's statements..

I don't get why this is so hard to grasp ... You seem to be talking thru the the prism of ACTUAL EXPERIENCE with a BLA Mod...madtheory is not..

THAT is the root of my issue with his stance in this thread..

I personally have no problem and have seen more then 1 person state that they were not satisfied with what they got back from their investment in the mod. Why would I or anyone have an issue with that?

It is arguing from a point of ignorance (because without actual usage,testing what other point could one possibly argue from?) that is objectionable and frankly seems trollish to me..

There is not a piece of gear on the planet that someone does not think is crap while someone else thinks is indespensible... that's a good thing..

How can you be credible/pertinent if you have no experience with said piece of gear?

Calling for proper A/B testing is a valid request...but it was/is? not the OP's request (at least as far as I can tell).That particular exercise is for another sub-forum on this board.

PS> I also wonder why it is somehow incumbent on BLA clients to somehow prove to sceptics that BLA mods are worth the value. I'm curious, how would we do that that? How would we be able to quantify what level (if any) improvement is deemed "worth the price" to you? It seems to me that if you really have a hard on to know if BLA is a scam or not YOU, not us should do the work because personally I already know no matter what you may or may not "discover" that I got the value I wanted for my investment.That might not be the case for you - that's your question to answer..

[Gearhero]

Thread wouldn't be complete w/out stacking theory. Also make sure you use pres with atleasr a 10000 uv slew rate and a snr 200 db below ambient noise. Please only test with with music covering Dark Side of the Mood. Otherwise differences won't be as appearant. Also realize those of us without BLA or Lavry Gold DAs won't be able to hear the huge subleties that those with proper converters will hear.

[madtheory]

Gearhero, yes, I think you're on to something!

Originally Posted by D K

It is arguing from a point of ignorance (because without actual usage,testing what other point could one possibly argue from?) that is objectionable and frankly seems trollish to me.

OK, here's an example, just to illustrate the point. There is a ton of mics out there that state they're flat from 20 Hz to 20kHz with ideal self noise, yet one is able to discern which ones are worth actually taking the time to test by listening and recording, and which ones are not worth the trouble. How does one make such a judgement? By being discerning. There is a certain amount one can assume from the given (or not given!) specs and the sales pitch. Discernment comes from being informed about the technicalities involved. One does not need to be able to design and build a mic to make such a judgement. Just solid, basic audio electronics knowledge, a.k.a. science. How does one acquire this knowledge? Not from the internet, but from good books, and maybe a little course or two at a local college.

I've been getting encouraging emails from those in agreement with me, so I will take some time to go through all the tech specs and sales pitches I have to hand and point out the questionable science. I will include references to reliable books and articles that explain the science. This will easily take a week to do well. It should make things easier for those who are avoiding doing their own research.

[mobius.media]

Originally Posted by Gearhero

Thread wouldn't be complete w/out stacking theory.

It's not a theory. It's mathematical fact, and it was demonstrated as best as was possible, experimentally.

It is also fact mathematically that artifacts from conversion will be amplified by typical mixing/mastering processes, and thus a pure loopback, while useful, is not the same as a comparison involving equal mix processing applied to each loopback (as in an actual mix).

And it is also almost certainly true that ABX clip tests, while also useful, do not replace long hours, days, and months working with gear while putting it through its paces it varied applications.

[madtheory]

mobius, he was kidding... but at least we can agree that a loopback is useful. Let's do those first, it's an easy test to do, and we'll discuss the other contentious issues later. OK?

[Budda]

Originally Posted by madtheory

Gearhero, yes, I think you're on to something!

OK, here's an example, just to illustrate the point. There is a ton of mics out there that state they're flat from 20 Hz to 20kHz with ideal self noise, yet one is able to discern which ones are worth actually taking the time to test by listening and recording, and which ones are not worth the trouble. How does one make such a judgement? By being discerning. There is a certain amount one can assume from the given (or not given!) specs and the sales pitch. Discernment comes from being informed about the technicalities involved. One does not need to be able to design and build a mic to make such a judgement. Just solid, basic audio electronics knowledge, a.k.a. science. How does one acquire this knowledge? Not from the internet, but from good books, and maybe a little course or two at a local college.

I've been getting encouraging emails from those in agreement with me, so I will take some time to go through all the tech specs and sales pitches I have to hand and point out the questionable science. I will include references to reliable books and articles that explain the science. This will easily take a week to do well. It should make things easier for those who are avoiding doing their own research.

Must not be busy with the mixing, if you have the time and desire to summarize other people's shortcomings.

I have a Phoenix Audio DRS-1. I got a frequency response plot with it that says it's pretty darn near linear in the audio spectrum. I don't know if it's ACTUALLY flat, but I know it sounds different than my Hardys - which, I believe, are purportedly pretty flat too.

Oh, and before I knew better, I tried to mix on M-Audio BX-5's. They too came with their own frequency response plots that seemed to indicate that they were relatively flat. I liked how those graphs looked - I quickly came to greatly dislike how the speakers actually sounded.

Probably for the third time in this thread alone, I will pick on something like a 1073. Even Mr. Neve himself says that they were simply the best he could do at the time, that his new RND stuff specs better and should perform better. THE DESIGNER HIMSELF! Still, people like his old stuff - I know I do - because of how useful his tools are in the recording process, specs be damned.

No one in this thread cares what you say because you haven't used the products in question. If you want to start your own thread, go ahead. You are serving no other purpose here.

People find the BLA stuff useful - specs be damned. Some have said it wasn't worth it, but no one who's actually used it seems to think that their box sounds the same or worse, but universally better.

[mobius.media]

Originally Posted by madtheory

at least we can agree that a loopback is useful. Let's do those first, it's an easy test to do, and we'll discuss the other contentious issues later. OK?

I don't think there's anything contentious about anything I said. If you or anyone else has reason to disagree, I'd be curious to hear it.

[madtheory]

The many lengthy threads on stacking theory would certainly indicate how contentious the idea is.

Originally Posted by Budda

Must not be busy with the mixing, if you have the time and desire to summarize other people's shortcomings.

Pot calling the kettle black? Definitely. Incidentally, I actually prefer this to vegging in front of the TV after work- you might make yourself aware of timezones. As for your examples, once again you cannot see the wood for the trees. Incidentally, I like the 1073 too, but you must admit that the colour of it is not always appropriate.

[mobius.media]

Originally Posted by madtheory

The many lengthy threads on stacking theory would certainly indicate how contentious the idea is.

All the threads in the world won't change the reality that it is mathematical fact. Unless you know some math I don't. If that is the case, I hope you will share it.

[madtheory]

Originally Posted by mobius.media

All the threads in the world won't change the reality that it is mathematical fact. Unless you know some math I don't. If that is the case, I hope you will share it.

Possibly correct, but their quantity and length shows that it is, as I said, contentious:
http://dictionary.reference.com/browse/contentious
Feel free to start another stacking thread, I won't be debating that one, but you should stick to the point, it helps.

[zacheus83]

Wouldn't it be simpler to record a session from a separate digital multitrack into an unmodded and modded unit? The session would be fixed (as opposed to recording live). We'd have to slightly take into account the original daw, but aren't we really just comparing the differences between modded and unmodded unit? If stacking is real or myth won't matter in this case if we recorded 16+ tracks from the original session into both units. This would take 2 passes per unit (8 I/O each pass) and would only take 20 minutes. This would require 2 laptops and a 3rd digital multitrack (computer or VS2480, etc..., so that the modded/unmodded units D-A doesn't factor in). Wouldn't this be the simplest way to compare differences by ear?

[mobius.media]

Originally Posted by madtheory

Feel free to start another stacking thread, I won't be debating that one, but you should stick to the point, it helps.

If we are talking about discerning differences between pieces of gear, and the limitations of possible test protocols to do so, it is perfectly on topic.

I have no interest in starting a new thread. As far as the math is concerned it is settled fact.

Originally Posted by zacheus83

Wouldn't it be simpler to record a session from a separate digital multitrack into an unmodded and modded unit? The session would be fixed (as opposed to recording live). We'd have to slightly take into account the original daw, but aren't we really just comparing the differences between modded and unmodded unit? If stacking is real or myth won't matter in this case if we recorded 16+ tracks from the original session into both units. This would take 2 passes per unit (8 I/O each pass) and would only take 20 minutes. This would require 2 laptops and a 3rd digital multitrack (computer or VS2480, etc..., so that the modded/unmodded units D-A doesn't factor in). Wouldn't this be the simplest way to compare differences by ear?

You don't need to do anything like that. All you need is 8 passive splitters out of 8 preamps. Send one set to one DAW, and another to the other DAW (eg. a laptop). Record simultaneously.

If you want to be serious, you can then add plugins set equally on both multitracks to see how a simple mix would come together. This is how Kenny did it in his test.

An 8 channel mix is not a 96 channel mix, but it is better than nothing.

A simple mix loopback should also be done. And lastly, an RMAA spec analysis (freeware test app) is needed.

And as stated before, none of this replaces the value of days, weeks, and months working with a piece of gear in varied application.

[madtheory]

mobius, we're actually talking about whether stacking theory is contentious, not if it's a valid theory...

Yes a loopback is even easier. Clock jitter is eliminated because AD and DA are in sync, but that's fine because jitter is contentious, and the clock mod is optional. Given the sales pitch about the other mods, I'm sure it will reveal differences that should be audible. Furthermore few loopback tests would be a good start, we need action, the words are getting tiresomely distracting.

[mattyblue]

Originally Posted by madtheory;576[LIST=1

[/LIST]8429]mobius, we're actually talking about whether stacking theory is contentious, not if it's a valid theory...

A loopback is even easier. Clock jitter is eliminated because AD and DA are in sync, but that's fine because jitter is contentious, and the clock mod is optional. Given the sales pitch about the other mods, I'm sure it will reveal differences that should be audible. Furthermore few loopback tests would be a good start, we need action, the words are getting tiresomely distracting.

Clock jitter is always going to be there inherently within the internal clock. You would definitely here a difference between the internal clock and the one bla puts in the 002/003, being that the internal has about 20ns of jitters and the bla has about 3-5 ps of jitter

[madtheory]

No, if you loopback it means you are using a DA and an AD that are on the same PCB, so they have the same clock. Therefore, the jitter is eliminated.

[mattyblue]

Originally Posted by madtheory

No, if you loopback it means you are using a DA and an AD that are on the same PCB, so they have the same clock. Therefore, the jitter is eliminated.

If you send a signal out of the d/a and back into the a/d the same clock that they are sharing has its own jitter that is there from the beginning. You can't eliminate jitter that is already coming from the clock signal

[madtheory]

Jitter means variations in the sampling rate, right? If it speeds up on the DA, it speeds up identically on the AD so it makes no difference to the captured audio.

[Budda]

Originally Posted by madtheory

The many lengthy threads on stacking theory would certainly indicate how contentious the idea is.

Pot calling the kettle black? Definitely. Incidentally, I actually prefer this to vegging in front of the TV after work- you might make yourself aware of timezones. As for your examples, once again you cannot see the wood for the trees. Incidentally, I like the 1073 too, but you must admit that the colour of it is not always appropriate.

Buddy, the only continents I haven't been to are Australia and Africa. I consider Antarctica a nice goal to get to, but completely optional. Nice to know you understand that it is currently around 7pm local time, and I'm taking a break between the day's activities and a mastering session that will start at 8. No time today to veg in front of the TV - just a little break over dinner. And maybe over lunch for the other posts. If I were in central Europe, or maybe Argentina, I might be able to post around 1 during siesta, or around 4 when people have a snack (do you have tea where you are?), and then maybe around dinnertime at 9. Yes, I'm familiar with time zones.

As for wood for the trees - buddy, you misquote the Neve statement - or at least, remove the context of my comment - and tell me I don't see the big picture. Irony indeed.

But I'm not going to get further into the personal thing. I'm not the brightest guy around, so I'm not going to tell everyone that my version of reality is the correct one - I just believe in it strongly. You do your own reality. Reconcile yourself to the fact that it may not be someone else's version and leave it at that.

I've tried to be civil. I've tried to contribute to the thread as an owner as the O/P requested, and I've tried to defend guys who I think are doing decent work that I personally find value in (BLA) because they were being slagged by know-it-alls who have never used their gear. I picked this fight with peeder/spicemix/whatever the third name was, both on the DUC and here, and both times I got fed up and gave up. Then again, both times he was kicked out of the discussion by the mods - gather your own conclusions from that.

Good luck to all of you. I'm off to master.

LOVES ME MY BLA!

[madtheory]

Wow... OK, one last try. I'll spell it out for you. With regard to your attempt at an insult, it was 11 pm when I posted, while the adverts were on during "The Genius of Charles Darwin" presented by Richard Dawkins.

So, seriously, do you often find yourself working at 11 pm?

[mattyblue]

Originally Posted by madtheory

Jitter means variations in the sampling rate, right? If it speeds up on the DA, it speeds up identically on the AD so it makes no difference to the captured audio.

Jitter is random errors within the clock. So although they receive the same clock they still have the same amount of jitter and you can not cancel out jitter. i.e, if a clock has 250 ps of jitter then that means that both a/d and d/a are receiving that clock signal with the same amount of jitter. So although you won't be adding jitter to the amount of inherent jitter in the overall audio signal will come out to be 250 ps give or take a little.

[madtheory]

To be precise it's not the same amount of jitter, it's the same jitter. If you have an ADAT deck you can test this. Set up the loopback, but clock your interface from the ADAT deck. While you record the loopback, randomly change the sampling rate up and down, by using the pitch control on the ADAT. The deck will show you what the sampling rate is on its display. Go to town on it, make those pitch changes wild. When you've completed the recording, set the clock on your interface to internal. Play back the recording. If it is indeed as you describe, the pitch variations will be played back. However, I think you will find that the pitch does not change at all.

By the way, jitter is not necessarily random, for example a poor PSU can put the AC on the clock line, giving periodic jitter. However, jitter is always undesirable. You really need to do the research I described earlier...

[mattyblue]

Originally Posted by madtheory

To be precise it's not the same amount of jitter, it's the same jitter. If you have an ADAT deck you can test this. Set up the loopback, but clock your interface from the ADAT deck. While you record the loopback, randomly change the sampling rate up and down, by using the pitch control on the ADAT. The deck will show you what the sampling rate is on its display. Go to town on it, make those pitch changes wild. When you've completed the recording, set the clock on your interface to internal. Play back the recording. If it is indeed as you describe, the pitch variations will be played back. However, I think you will find that the pitch does not change at all.

By the way, jitter is not necessarily random, for example a poor PSU can put the AC on the clock line, giving periodic jitter. However, jitter is always undesirable. You really need to do the research I described earlier...

I think we are talking about 2 different kinds of jittter. I'm more referring to the jitter that is inherent at the actual clock and what gets recorded to the actual data stream. If I record a piece of audio, the timing of that clock is embedded in the converted data. The timing is going to be constantly the same no matter what. Yes it is true that there can be some anticipated jitter, but variables aside, jitter that is inherently in the actual crystal oscillator itself is random and you can not cancel that out. That was my point.

[madtheory]

Right, so do you see that if we do a loopback on a Digidesign 192, or any of the modified boxes under discussion, that the AD and DA run from the same crystal based oscillator? Each converter clocks a value into and out of its buffer at precisely the same voltage and time. So the captured audio will not show any jitter added during the loopback. This is a component level synchronisation, on the actual PCB. It's totally different to trying to synchronise two different crystal based oscillators, as you would be when the DA and AD are physically separate. The common clock is the key. In the ADAT clocked test, the ADAT is the common clock.

[mattyblue]

Originally Posted by madtheory

Right, so do you see that if we do a loopback on a Digidesign 192, or any of the modified boxes under discussion, that the AD and DA run from the same crystal based oscillator? Sso the captured audio will not show any jitter added during the loopback.

Right that was my point that you are strictly seeing the conversion as it is (flaws and errors as well) and nothing is artificially being added to skew the test. But there will always be the jitter that does come straight from the clock which is part of the conversion process and which will ultimately be put to the test.

[madtheory]

Originally Posted by mattyblue

Right that was my point that you are strictly seeing the conversion as it is (flaws and errors as well) and nothing is artificially being added to skew the test. But there will always be the jitter that does come straight from the clock which is part of the conversion process and which will ultimately be put to the test.

Of course there may be jitter in the clock, but the point is if DA and AD are run from the same oscillator, the jitter is not imposed on the audio in a loopback. Each converter tracks the other.

Also, please note that I've added to my explanation after you quoted it.